This paper reports on findings obtained in a larger research project that examines the effect of several functional foods supplied through chewing gums on caries prevention in an adult population. This paper was focused on the effect of low-dose xylitol.
Ethical approval
The present study was carried out in Sassari (Italy) under the supervision of the WHO Collaborating Centre for Epidemiology and Community Dentistry of Milan, Italy, and lasted from September 2012 to June 2015. The study was designed as a randomized clinical trial, approved by the Ethics Committee of the University of Sassari (n°1083/L 23/07/2012), and registered (Protocol Registration Receipt NCT02310308) at http://www.clinicaltrial.gov.
Study population
Data from the Italian National Institute for Statistics for 2011 gave the number of 30–45-year olds living in the town of Sassari as 22,614.
The inclusion criteria were as follows: age range between 30 and 45 years; presence of a minimum of 12 natural teeth; presence of at least one cavitated (D2/3) caries lesion, but not more than three; a salivary concentration of MS equal or upper to 105 CFU/mL saliva; no current periodontitis (no sites of probing pocket depth ≥5 mm or attachment loss of ≥2 mm, apart from gingival recession); absence of dysfunction of temporomandibular joint; good health systemically as assessed by a medical questionnaire; no allergy to any of the ingredients of the study products; no orthodontic banding or removable prosthesis; and no use of antibiotics or participation in a clinical study in the previous 30 days. The use of antibiotics was also recorded at interim examinations (6, 12, and 24 months) before saliva sampling, and subjects using antibiotics 30 days before the evaluation were excluded.
Sample size for preliminary screening was performed through G*Power 3.1.3 for Apple using logistic regression with an odds ratio of 1.8 and an error probability of 0.04; the total sample was set at 312.
In order to get statistical comparable results, the number of subjects per group to be included in the analysis was calculated. Considering a 40% difference among groups to be significant and a 95% probability of obtaining a significant difference between groups at the 5% severity, the resulting number of subjects per group was set in 64.
With the collaboration of the municipal electoral registry office, a letter explaining the purpose of the study and the informed consent were randomly distributed to 5% (1131 subjects) of the age group considered living in Sassari. A total of 480 subjects (42.4% acceptance rate) accepted to participate and were examined for conditions that would preclude participation. The flow chart, displayed in Fig. 1, shows the design of the study.
Randomization was performed (GCampus) using Excel 2014 in permuted blocks of two or four with random variation of the blocking number, and two groups were created: (1) the first group received gums containing the same mixture of polyols except xylitol; 2) the second group received gums containing the same polyols mixture plus a low amount of Xylitol. The clinical examination was repeated at the end of the chewing gum administration period (12 months) and at the end of the experimental period (24 months).
A benchmark examiner (GCampus) trained and calibrated one examiner (GCarta) that performed all dental screenings. Baseline training consisted in 1-day (6 h) theoretical course, followed by examination of 54 extracted teeth plus a session of 120 photographs of extracted teeth. Two days after the theoretical course, a clinical training involving examination of 55 adults was performed. The subjects were re-examined after 72 h. Inter-examiner reliability with the “benchmark” (GCampus) was evaluated using fixed-effect analysis of variance. Intra-examiner reproducibility was assessed as the percentage of agreement using Cohen’s kappa statistic [17]. Good reliability was found between examiner and benchmark (p = 0.15) with a low mean square of error (0.47). Intra-examiner reliability was also high (Cohen’s kappa = 0.88). Interim and follow-up training was also performed. At the interim clinical evaluation (12 months), 47 adults not enrolled into the trial were re-examined after 72 h with a good reliability between examiner and benchmark (p = 0.14) with a mean square of error (0.49) and a high intra-examiner reliability (Cohen’s kappa = 0.91). Before the final follow-up examination (24 months), 45 adults not enrolled into the trial were re-examined after 72 h with an inter-examiner reliability (examiner vs benchmark (p = 0.14) with a mean square of error (0.49) and a high intra-examiner reliability (Cohen’s kappa = 0.91). Subjects were examined using a mouth mirror and a Community Periodontal Index probe (approved by the WHO) under optimal lighting. The ICDAS (International Caries Detection and Assessment System) index was used to register caries at tooth level as initial or moderate or extensive lesions and the number of filled and missing teeth for caries [18]. Initial caries lesion can be defined as a primary lesion, which has not reached the stage of an established lesion with cavitation. Moderate caries lesions are defined as white or brown spot lesion with localized enamel breakdown or an underlying dentine shadow without visible dentine exposure. Severe caries lesions are defined as distinct cavity in opaque or discoloured enamel with visible dentine [19]. Those participants who referred to consume more than three pieces of sugar-free chewing gum a day were excluded. The elected participants agreed not to consume any other chewing gums than those supplied for the study.
All participants were residents in an area with a low natural fluoride content in the drinking water (0.04 mg/L) (http://www.abbanoa.it/distretto-6-sassari1), but they reported to use a fluoridated toothpaste on a regular basis.
Microbiological evaluation
Immediately after the clinical assessment, an evaluation of MS concentration in saliva was performed. Nonstimulated whole saliva was collected over 150 s in sterile vials (Nunc, Kamstrup, Denmark). The samples were transported to the Department of Microbiology and processed within 45 min after collection. The samples were serially diluted in sterile PBS (Sigma Chemicals, St. Louis, MO, USA). Aliquots of 5 μL were inoculated on mitis salivarius bacitracin agar, a medium that at concentrations of bacteria 1 × 103 to 1 × 1010/mL shows a good sensitivity and selectivity in MS detection. The plates were incubated in a 5% CO2 environment at 37 °C for 72 h after which the colony-forming units were identified by morphology, size, and colour and counted in a stereomicroscope.
Plaque pH measurements
Inter-proximal plaque pH of each subject was evaluated using pH indicator strips [20], which measure a pH value in the range 4.0–7.0 (Spezialindikator, pH range 4.0–7.0; Merck, Darmstadt, Germany), with a resolution of 0.2–0.5 pH unit; in addition, the strips are easy to use. The strips were cut into four pieces (approx. 2 mm in width) in order to be more easily inserted into the inter-proximal space and held into the inter-dental space for 10 s, after which they were removed and their colour compared to the colour index scheme supplied by the manufacturer. The pH was determined to one decimal of the value.
Plaque pH was assessed at baseline, after 6 and 12 months of chewing gum use and 12 months after the cessation of chewing gum use. For each subject, three measurements were carried out on two sites, between the second premolar and the first molar, right and left of the upper jaw; the average pH value was later calculated. Recordings were performed before and at 2, 5, 10, 15, 20, and 30 min after a mouth rinse with 10% sucrose and carried out by one examiner (FC). Area under the curve (AUC), described as the area between reference pH (6.2 or 5.7) line and the pH curve, was calculated using a computer-based program [21]. The area under the curve at pH 5.7 and 6.2 (AUC5.7 and AUC6.2) was used as a reference for the dissolution of enamel and dentine, respectively.
Treatment and sample collection
Overall, 480 subjects were examined, 331 fulfilled the inclusion criteria, and 179 accepted to be enrolled in the trial.
All chewing gums were produced and supplied by Perfetti Van Melle SpA (Lainate, Italy). The polyol chewing gum was sugar-free containing 28% isomalt, 31% sorbitol, 9% mannitol, and 1% maltitol syrup. Xylitol chewing-gum contained 30% of Xylitol, 26% Sorbitol, 11% Mannitol and 1% Maltitol syrup.
All chewing gums weighed 1.4 g each and were identical in colour, shape, and taste. Chewing gums were supplied in plain white containers coded as ‘green’ or ‘blue’ according to the group. The code was sealed by an independent monitor and not broken until the statistical analysis was finalized.
The subjects were instructed to chew for 5 min two pellets in the morning, two after the midday meal, and one in the afternoon; the total daily intake of xylitol was 2.5 g/day. The subjects were asked to make no changes in their dietary and oral hygiene habits. Tooth brushing was not allowed for at least 1 h after the use of chewing gums. All subjects received a fluoridated toothpaste containing 1450 ppm NaF (Mentadent P; Unilever Italia, Milan, Italy) to be used during the experimental period. They were also asked to avoid any other oral hygiene adjuvant and any commercial xylitol or sorbitol product throughout the study period. The body’s tolerance to different polyols was assessed by means of a questionnaire administered to the participants shortly after the gum distribution had started and 6 months later, while the study was still proceeding. The questions focused on the potential side effects of using the gum. In order to evaluate the success of the chewing gum intake, participants were given chewing gums necessary for a single month at a time and asked to return the empty packs when receiving those for the following month.
Statistical analysis
The tooth as the unit of analysis was evaluated as follows: first, the net caries increment for initial (ICDAS 1 and 2), moderate (ICDAS 3 and 4), and extensive (ICDAS 5 and 6) caries severity using ICDAS (Δ-initial, Δ-moderate, and Δ-extensive) was calculated. The number of events is the sum of the Δ-caries change of status recorded at baseline, 12 months of gum use, and after 12 months of no-gum use. Secondly, caries experience expressed as the sum of extensive caries lesions plus the number of filled and extracted teeth due to caries was calculated and consequently the caries increment was recorded. Differences across mean number of events between groups for each variable were evaluated using the nonparametric Mann–Whitney U test.
The data on inter-proximal plaque pH at baseline, 6 months of gum use, 12 months of gum use, and after 12 months of no-gum use were analysed for statistically significant differences using repeated measures of ANOVA.
Differences in proportion relating to microbiological counts at baseline and follow-ups were assessed using equality of proportion test. The lowness curve was used to describe the trend of plaque pH and salivary MS. The effectiveness of the treatment was assessed for those who fully followed the protocol (per-protocol subjects) by calculating the reduction in risk ratio (RRR) and the related number needed to treat (NNT) value [22]. An event was defined as the change of status at tooth level, i.e. the development of a new lesion or the progression of an existing lesion to a more severe stage. All data were analysed using the software STATA® (v13 for Mac). For all statistical analyses, the statistical significance was set at α = 0.05.